It's alive! Or at least my DC power supply is alive. 25.2 volts on both sides, within .01 volts of each other. The future front panel LEDs are the brightness I wanted (I mocked up with another power supply with the resistor/LED pairing), so now I will install those, now to get those power supply boards stuffed and assembled.
Been dragging my butt on this, but excited to have this milestone passed!
Been dragging my butt on this, but excited to have this milestone passed!
Hello all,
I built an Aleph J with single-ended inputs using the BrianGT boards. In order to simplify the build I omitted R3 and C1 and there was no connection to IN-.
I recently decided that I wanted to try balanced connections end-to-end in my system, so I ordered a pair of Neutrik XLR jacks, 1uF caps (C1) and 22.1K resistors (R3) and installed them in the J. I disconnected the twisted pairs from the RCA jacks, wired them to pins 2 & 3 on the XLR jacks, and wired pin 1 to one of the jack's mounting screws. On the board end, I left the + wire in IN+, and moved the - wire from signal ground to IN-. It follows that there is now no wire connected to the signal ground. I have confirmed 0 ohms from pin 1 to chassis ground, and from signal ground to common ground, on both channels.
Despite being unable to convince myself, I suspect that I am missing a connection, because there is hardly any output to the speakers (faint sound at max volume) and back on the bench I can't get a stable bias or DC offset with a shorting plug in the input. Temporarily reinstalling the RCA jacks (now with a jumper from signal ground to IN-) allows me to once again set the bias and DC offset accurately.
What am I overlooking?
I built an Aleph J with single-ended inputs using the BrianGT boards. In order to simplify the build I omitted R3 and C1 and there was no connection to IN-.
I recently decided that I wanted to try balanced connections end-to-end in my system, so I ordered a pair of Neutrik XLR jacks, 1uF caps (C1) and 22.1K resistors (R3) and installed them in the J. I disconnected the twisted pairs from the RCA jacks, wired them to pins 2 & 3 on the XLR jacks, and wired pin 1 to one of the jack's mounting screws. On the board end, I left the + wire in IN+, and moved the - wire from signal ground to IN-. It follows that there is now no wire connected to the signal ground. I have confirmed 0 ohms from pin 1 to chassis ground, and from signal ground to common ground, on both channels.
Despite being unable to convince myself, I suspect that I am missing a connection, because there is hardly any output to the speakers (faint sound at max volume) and back on the bench I can't get a stable bias or DC offset with a shorting plug in the input. Temporarily reinstalling the RCA jacks (now with a jumper from signal ground to IN-) allows me to once again set the bias and DC offset accurately.
What am I overlooking?
It might be worth noting that the input to the preamp was single-ended while the output was balanced - I don't yet have enough balanced cables to run end-to-end yet.
XLR pin 1 to board Ground, XLR pin 2 to board IN+, XLR pin 3 to board IN-
Edit: What resistance do you measure from XLR pin 1 to board Ground?
Edit: What resistance do you measure from XLR pin 1 to board Ground?
Last edited:
Thanks, Ben and Kevin. I will try it with the direct connection from pin 1 to the bosrd and report back.
Haven't tried the balanced connection again yet, but unbalanced still works so at least I haven't broken anything. 
Ben Mah - to answer your question, 0 ohms between pin 1 and signal ground.
Ben Mah - to answer your question, 0 ohms between pin 1 and signal ground.
With pin 1 wired to signal ground, I can bias the amp and adjust DC offset to 0.
With any luck the pair of cables intended to connect my preamp to the Aleph J will arrive on my doorstep tomorrow.
With any luck the pair of cables intended to connect my preamp to the Aleph J will arrive on my doorstep tomorrow.
Does anyone has any experience using ~+/-30V rails for the Alef J? Implemented using 22V 500 VA toroid PSU.
I also managed to obtain a pair of LSJ689, which has a higher max V capacity but lower transconductance or should I use the original LSJ74s that have lower V max?
I also managed to obtain a pair of LSJ689, which has a higher max V capacity but lower transconductance or should I use the original LSJ74s that have lower V max?
ZM is Chicken
also being first Chicken in the history, having Cascode Fetish
contrary to all other Greedy Boyz - having Fets Fetish, mine is not Fet one, even if mine preserving Fets of being Torched and declared Maids of Devil
in short - do what you want, but don't cry to me later, asking where to buy JFets and how old ones are expensive and new ones are expensive and nogood
also being first Chicken in the history, having Cascode Fetish
contrary to all other Greedy Boyz - having Fets Fetish, mine is not Fet one, even if mine preserving Fets of being Torched and declared Maids of Devil
in short - do what you want, but don't cry to me later, asking where to buy JFets and how old ones are expensive and new ones are expensive and nogood
Nelson has mentioned on a few occasions that, if one is to push the JFETs hard, the ones he tested (Toshibas) worked surprisingly well. So, with Toshiba JFETs (available here in the store), + and - 28V DC could be an option for the brave ones.
I have + and - 24.5 V DC on my Aleph J with LSJ JFETs - so far so good. That voltage varies between 24 and 25V DC, depending on the mains supply (240 to even 250 V AC .. sometimes 😱 ).
But, the real question is... what are you going/trying to achieve with that voltage? More power? This ain't gonna happen - especially with a 4-ohm load, where the max output current is set and depends on - the quiescent current.
I am not sure what would happen with an 8-ohm load and higher supply voltage...? The heat dissipation will increase (for the same, given Iq) <- that is for sure.
So, the best thing to do for starters is... decide what power dissipation are you happy with (you would tolerate as a safe dissipation <- this depends on your build and room temperature), per each MOSFET. I would not go above 25W per each MOSFET.... but that's me. Which in turn means.... you've guessed it: + and - 24V DC supply and Iq of approximately 25W power dissipation per each MOSFET (very safe).
I have + and - 24.5 V DC on my Aleph J with LSJ JFETs - so far so good. That voltage varies between 24 and 25V DC, depending on the mains supply (240 to even 250 V AC .. sometimes 😱 ).
But, the real question is... what are you going/trying to achieve with that voltage? More power? This ain't gonna happen - especially with a 4-ohm load, where the max output current is set and depends on - the quiescent current.
I am not sure what would happen with an 8-ohm load and higher supply voltage...? The heat dissipation will increase (for the same, given Iq) <- that is for sure.
So, the best thing to do for starters is... decide what power dissipation are you happy with (you would tolerate as a safe dissipation <- this depends on your build and room temperature), per each MOSFET. I would not go above 25W per each MOSFET.... but that's me. Which in turn means.... you've guessed it: + and - 24V DC supply and Iq of approximately 25W power dissipation per each MOSFET (very safe).
It turns out that, in my amp, pin 1 must be wired to signal ground in order for it to work.
Connecting pin 1 only to the chassis via one of the jack's mounting screws does NOT work - in my amp.
Connecting pin 1 only to the chassis via one of the jack's mounting screws does NOT work - in my amp.
it works (in fact, in both ways) in amps where signal ground is connected, in some (proper) way, to chassis
so, if that detail is still missing in your amp, it isn't proper
is there at least 10-15R NTC between audio GND and chassis (which must be connected to safety GND)?
except in case of being made of wood or plastic or any other non-conducting material
so, if that detail is still missing in your amp, it isn't proper
is there at least 10-15R NTC between audio GND and chassis (which must be connected to safety GND)?
except in case of being made of wood or plastic or any other non-conducting material
Chassis is connected directly to safety ground.
PS ground is connected to chassis through NTC 22.
Board common ground is connected to PS ground (i.e. reference for V+/V-).
Edit: as reported to Ben Wah previously, meter reads 0.00 ohms between signal ground and chassis.
PS ground is connected to chassis through NTC 22.
Board common ground is connected to PS ground (i.e. reference for V+/V-).
Edit: as reported to Ben Wah previously, meter reads 0.00 ohms between signal ground and chassis.
in that case, it should work with pin 1 connected directly/only to chassis, when using balanced input signal
though, which way will be less/no hum, depends of other things in your system, not just of amp
anyway, it doesn't matter, what is - is that you found arrangement working for you
though, which way will be less/no hum, depends of other things in your system, not just of amp
anyway, it doesn't matter, what is - is that you found arrangement working for you
Shouldn't the ohmic value of the NTC be the DMM reading measuring signal gnd to chassis?